Evolution of Whirly1 in the angiosperms: sequence, splicing, and expression in a clade of early transitional mycoheterotrophic orchids

Abstract

The plastid-targeted transcription factor Whirly1 (WHY1) has been implicated in chloroplast biogenesis, plastid genome stability, and fungal defense response, which together represent characteristics of interest for the study of autotrophic losses across the angiosperms. While gene loss in the plastid and nuclear genomes has been well studied in mycoheterotrophic plants, the evolution of the molecular mechanisms impacting genome stability is completely unknown. Here, we characterize the evolution of WHY1 in four early transitional mycoheterotrophic orchid species in the genus Corallorhiza by synthesizing the results of phylogenetic, transcriptomic, and comparative genomic analyses with WHY1 genomic sequences sampled from 21 orders of angiosperms. We found an increased number of non-canonical WHY1 isoforms assembled from all but the greenest Corallorhiza species, including intron retention in some isoforms. Within Corallorhiza, phylotranscriptomic analyses revealed the presence of tissue-specific differential expression of WHY1 in only the most photosynthetically capable species and a coincident increase in the number of non-canonical WHY1 isoforms assembled from fully mycoheterotrophic species. Gene- and codon-level tests of WHY1 selective regimes did not infer significant signal of either relaxed selection or episodic diversifying selection in Corallorhiza but did so for relaxed selection in the late-stage full mycoheterotrophic orchids Epipogium aphyllum and Gastrodia elata. Additionally, nucleotide substitutions that most likely impact the function of WHY1, such as nonsense mutations, were only observed in late-stage mycoheterotrophs. We propose that our findings suggest that splicing and expression changes may precede the selective shifts we inferred for late-stage mycoheterotrophic species, which therefore does not support a primary role for WHY1 in the transition to mycoheterotrophy in the Orchidaceae. Taken together, this study provides the most comprehensive view of WHY1 evolution across the angiosperms to date.

Keywords: Corallorhiza; Whirly1; genomic stability; intron retention; mycoheterotrophy; orchid; plastome evolution; transcription factor.

Figure 1

Overview of Corallorhiza species included in this study, showing plastid and nuclear phylogenetic relationships, inflorescence, trophic status (fully vs. partially mycoheterotrophic), plastome size (bp), number of putatively functional plastid genes, and chlorophyll content (mean and standard deviation in nanograms of total chlorophylls per milligram of plant material). Phylogenetic, plastome, and chlorophyll content data are from Barrett et al. (2014).

Figure 2

Wrapped view of open reading frame-trimmed Corallorhiza WHY1 isoform alignment. Disagreements with the consensus sequence are highlighted. An ^* denotes a canonical isoform, while ^ψ denotes an isoform with a retained intron.

Figure 3

DNA alignment of canonical isoforms and Nanopore-sequenced genomic sequence from each Corallorhiza species with annotated WHY1 sequence from the Phalaenopsis equestris and Dendrobium catenatum reference genomes. Coding sequences for each reference genome are depicted by yellow annotations. Disagreements with the consensus sequence are shown in black. Percent identity of aligned sites is depicted as a histogram. An * denotes a canonical isoform, while ψ denotes an isoform with a retained intron.

Figure 4

DNA alignment of Corallorhiza isoforms assembled from each Corallorhiza species and Nanopore sequences for C. trifida, C. maculata, and C. striata with annotated WHY1 sequence from the Phalaenopsis equestris and Dendrobium catenatum reference genomes. WHY1 exons 1 and 2 are annotated in yellow and the locus ID tags for each reference genome are provided in the coding sequence (yellow) annotations. Disagreements with the consensus sequence are highlighted. Percent identity of aligned sites is depicted as a histogram. An ^* denotes a canonical isoform, while ^ψ denotes an isoform with a retained intron.

Figure 5

Detailed view of amino acid alignment of WHY1 sequences of orchid species. Note the phenylalanine substitutions unique to Corallorhiza wisteriana, C. maculata, and C. striata, relative to the other species sampled. The amino acid substitutions of interest are located at the 3’ portion of a region inferred to conform into an alpha helix structure (pink annotation) in Arabidopsis thaliana WHY1 (alignment row two).

Figure 6

Maximum likelihood inferred topology of WHY1 evolution with support values derived from 1,000 non-parametric bootstrap replicates. Lineages highlighted in red represent those for which statistically significant signal of episodic diversifying selection of WHY1 was detected by the adaptive branch-site random effects likelihood test.